- Server Data:
- Server Administrator/Contact: webmaster@soe.ucsc.edu
- Server Name: users.soe.ucsc.edu
- Server Port: 80
- Server Protocol: INCLUDED
- Virtual Host: users.soe.ucsc.edu
- Server Name: users.soe.ucsc.edu
- Request Data:
- User Agent/Browser: CCBot/1.0 (+http://www.commoncrawl.org/bot.html)
- Request Method: GET
- Remote Address: 38.107.191.82
- Remote Port: 35539
- Extra Path Info: /~amer4/axs/ax.pl
- Request Method: GET
Scalable Storage Systems
-
Scalable I/O Initiative
-
Preliminary stages of publication, but includes an incredible scientific workload survey.
http://www.cacr.caltech.edu/SIO/
-
Investigating issues in building storage systems capable of handling the
terabytes (and, soon petabytes) of data (MSS Research Group).
Exploring large scale information storage, retrieval
and visualization to develop distributed, scalable,
agent-based information management systems (CADIP).
MSS - http://www.cs.umbc.edu/~elm/Projects/mss.shtml
CADIP - http://www.csee.umbc.edu/cadip/
-
The High Performance Storage System, a collaboration between
IBM,
San Diego Supercomputer Center,
LLNL,
and several other laboratories and organisations.
http://www.sdsc.edu/hpss/
-
Hierarchical methods for massive scientific data visualization from UC Davis.
Part of the ASCI research efforts.
http://www.llnl.gov/asci-alliances/ap/hamann.html
-
Northwestern university's attempt to
"develop a scalable high-performance data-management and I/O system that provides
support for data management, query capability, high-performance accesses and automatic optimizations to large
datasets in hierarchical storage systems."
Part of the ASCI research efforts.
http://www.llnl.gov/asci-alliances/ap/choudhary.html
-
Scalable Availability from IBM Almaden.
http://www.almaden.ibm.com/cs/scalable.html
-
IBM Research on scalable multimedia/video servers.
http://www.research.ibm.com/vidservr/
-
Scaleable servers project at Microsoft's Bay Area Research Center (BARC).
Based in San Francisco and managed by Jim Gray.
http://research.microsoft.com/research/barc/Scaleable/
-
Scalable storage project from the University of Arizona. Related work includes
their contributions to the scaleable I/O initiative.
http://www.cs.arizona.edu/swarm/
-
A scalable, recoverable file system for AIX clusters.
http://www.research.ibm.com/caly/
-
Problem domain specific data and I/O management for large scale
scientific computing problems.
http://gizmo.lbl.gov/DM.html
-
Scalable Clusters at UCSD.
http://www-csag.ucsd.edu/projects/hpvm.html
-
The ScalaServer project considers a range of related issues in the design of
scalable, high-performance and highly available network servers.
http://www.cs.rice.edu/CS/Systems/ScalaServer/index.html
-
Princeton's Scalable High-performance Really Inexpensive Multi-Processor (SHRIMP)
project and the Parallel User-Level Server ARchitecture that provides a high
performance server from a cluster of commodity PCs.
PULSAR - http://www.cs.princeton.edu/shrimp/html/file_system.html
SHRIMP - http://www.cs.princeton.edu/shrimp/
-
Research to make PCs dance together fast, i.e. PC clustering software, and
a scalable distributed file system.
The Cooperative File System - http://discolab.rutgers.edu/projects/cfs/index.html
The Disco Group - http://discolab.rutgers.edu/
-
Vrije Universiteit project on wide area cluster computing.
http://www.cs.vu.nl/albatross/
-
The General Parallel File System (GPFS) is the product name for the
Tiger Shark file system developed at the
IBM Almaden Research Center.
Scalable cluster file system, originally developed for multimedia, but extended for
parallel computing applications.
http://www.almaden.ibm.com/cs/gpfs.html
-
The Stonybrook Video Server Project -- a LAN-based distributed video
server that provides end-to-end performance guarantees.
http://www.ecsl.cs.sunysb.edu/sbvs.html
-
A paper for the scalable I/O system TOPs from the University of Maryland.
ftp://ftp.cs.umd.edu/pub/papers/papers/ncstrl.umcp/CS-TR-3462/CS-TR-3462.ps.Z
Distributed File Systems
-
Coda
-
Mobile computing and disconnected operation.
http://www.coda.cs.cmu.edu/
-
The commercial offering from Transarc based on AFS technology.
http://www.transarc.com/Product/EFS/index.html
-
A distributed scalable file system. This page contains links to
a descriptive SOSP paper, and the corresponding powerpoint slides.
http://research.compaq.com/SRC/personal/thekkath/frangipani/home.html
-
University of Minnesota
Binary Operations Research Group B.O.R.G.
http://www.globalfilesystem.org/
-
University of Minnesota
Binary Operations Research Group B.O.R.G.
http://elbe.borg.umn.edu/
-
A real-time file system by Peter Bosch for Nemesis(Pegasus-II) and Linux.
http://www.huygens.org/~peterb/clockwise.html
-
Distributed systems research from Cambridge University.
http://www.cl.cam.ac.uk:80/Research/SRG/netos/old-projects/nemesis/
-
Vrije Universiteit project on the construction of very large wide-scale
distributed systems.
http://www.cs.vu.nl/~steen/globe/
-
The DISCWorld
file system prototype from Adelaide.
http://dhpc.adelaide.edu.au/projects/storage/
-
Exploring caching strategies for clients and servers in distributed file
system environments.
http://www.cs.usask.ca/homepages/faculty/bunt/projects/caching.html
-
A short retrospective on Berkeley's Sprite project.
http://www.cs.berkeley.edu/projects/sprite/retrospective.html
-
Brown University research in open distributed systems. Interoperability work.
http://www.cs.brown.edu/people/stp/summit/
-
Distributed OS from Vrije Univeriteit.
http://www.cs.vu.nl/pub/amoeba/
-
Large-Scale Distributed Systems research (BROADCAST) and
the Network of Excellence 21035 in
distributed computing systems architectures (Cabernet),
funded by
ESPRIT.
BROADCAST - http://www.newcastle.research.ec.org/broadcast/index.html
Cabernet - http://www.newcastle.research.ec.org/cabernet/index.html
Parallel I/O
-
MPI-IO
-
A Parallel File I/O interface for MPI.
http://hpcc.arc.nasa.gov/reports/annrep97/cas_arc/ww34.htm
http://www.llnl.gov/people/trj/goddard/
-
Excellent list of links to projects. Also includes
the parallel I/O
anecdotes.
http://www.cs.dartmouth.edu/pario/
-
The current Portable Parallel File System from the University of Illinois, Urbana-Champaigne.
Work in the context of the Pablo Research Group.
http://www-pablo.cs.uiuc.edu/Project/PPFS/PPFSII/PPFSIIOverview.htm
-
Parallel I/O for scientific applications at Lawrence Livermore National Laboratories.
WOrk that also collaborates with the Scalable I/O Initiative.
http://www.llnl.gov/sccd/lc/piop/
-
Algorithms for Transparent Parallel I/O from Duke University (TPIE),
and fast access to networked clusters of computers (Trapeze).
TPIE - http://www.cs.duke.edu/TPIE/
Trapeze - http://www.cs.duke.edu/ari/trapeze/
-
Scaleable storage through clusters of computers connected by high-speed networks, from
the University of California, Berkeley.
http://now.cs.berkeley.edu/index.html
-
A system for data-parallel computing with parallel data too large to
fit in main memory.
http://www.cs.dartmouth.edu/~thc/ViC/index.shtml
-
File system for parallel scientific workloads developed at Dartmouth, for use
on networks of workstations.
http://www.cs.dartmouth.edu/~nils/galley.html
-
Another parallel file system from Dartmouth.
http://www.cs.dartmouth.edu/~dfk/armada/
-
University of Maryland's high performance I/O and scientific computing effort.
http://www.cs.umd.edu/projects/hpsl/Chaos.htm
-
The Parallel I/O File System,
a commercial version of the Vesta system from IBM.
A file system that efficiently handles very large files.
http://www.rs6000.ibm.com/software/sp_products/piofs.html
-
Parallel-Storage-and-Processing Server (PS2), a high performance parallel storage
architecture, with applications to image storage and processing through the related
Gigaserver project.
http://diwww.epfl.ch/w3lsp/pub/research/ps2/ps2-intro.html
-
Finding algorithms to minimize the number of parallel disk accesses for large
scientific calculations.
http://www.cs.dartmouth.edu/research/html/cormen.html
-
Investigating parallel I/O at Clemson University.
http://parlweb.parl.clemson.edu/pvfs/
-
The scotch parallel storage testbeds at CMU for the investigation of
directions in RAID evolution.
http://www.pdl.cs.cmu.edu/PUBLICATIONS/HTML-Papers/Compcon95/final.fm.html
-
High performance I/O research for computational chemistry applications.
http://www-fp.mcs.anl.gov/chemio/
-
Low cost commodity parallel clsuters from Caltech.
http://www.cacr.caltech.edu/beowulf/index.html
Storage in Mobile Computing
-
Coda
-
Mobile computing and disconnected operation.
http://www.coda.cs.cmu.edu/
-
Mobile computing. Predictive file hoarding for disconnected operation in mobile environments.
Part of the TRAVLER project at UCLA.
http://fmg-www.cs.ucla.edu/geoff/seer.html
-
A mobile computing project at UCLA, including sub-projects on predicitve file hoarding,
predictive web prefetching, and file replication.
http://fmg-www.cs.ucla.edu/travler98/welcome.html
-
Decentralized consistency management for replicated data in mobile and
unreliable network environments.
University of Maryland.
http://www.cs.umd.edu/projects/deno/
-
Conserving power through the use of machine learning techniques to
predict the best disk spin-down policy
in mobile computers.
http://csl.cse.ucsc.edu/projects/power-conservation/
-
File caching support for mobile systems with network links of
variable quality.
http://www.cs.usask.ca/homepages/faculty/bunt/projects/mobile.html
-
European project covering OS support for multimedia, mobile computing, and security.
This is the umbrella project for
Pegasus II,
Broadcast,
CaberNet,
and Moby Dick.
http://www.huygens.org/
Workloads, Simulation and Benchmarking
-
DFSTrace
-
A flexible system tracing tool, capable of reading multiple trace formats.
http://csl.cse.ucsc.edu/projects/DFSTrace/
-
I/O pattern classification work from the University of Illinois, Urbana-Champaigne.
Work in the context of the Pablo Research Group.
http://www-pablo.cs.uiuc.edu/Project/IO/IOOverview.htm
-
File system workload characterization from Dartmouth.
http://www.cs.dartmouth.edu/~dfk/charisma/
-
A good disk simulation environement by Greg Ganger, Bruce Worthington and Yale Patt.
http://www.ece.cmu.edu/~ganger/disksim/
-
A complete machine simulation from Stanford.
http://simos.stanford.edu/
-
File system traces by Drew Roselli.
http://dawn7.cs.berkeley.edu:8080/
-
http://now.cs.berkeley.edu/Xfs/SpriteTraces/
-
File system traces by Geoff Kuenning.
http://fmg-www.cs.ucla.edu/geoff/seer_traces.html
-
University of Maryland CP's --
The Mambo
(Maryland Applications for Measurement and Benchmarking of I/O
on Parallel Computers)
http://www.cs.umd.edu/projects/hpsl/mambo/index.html
-
University of Kentucky --
Flie system traces by Randy Appleton
at the University of Kentucky.
Information and e-mail addresses only.
http://clarinet.dcs.uky.edu/~mbfs/traces.html
-
Performance and benchmarking of mass storage I/O, including fragmentation issues.
http://www-mount.ee.umn.edu/~okeefe/io_group/io.html
-
A CMU project for automatically characterizing disk performance
such as mechanical delays and on-board memory management algorithms.
An alternative to manufacturer-supplied details DIXtrac running under
Linux attempts to deduce over 100 performance critical parameters.
http://www.ece.cmu.edu:80/~schindjr/dixtrac/
-
Vino operating system from Harvard, including file systems research.
Analysis of file system performance and web caching.
http://www.cs.harvard.edu/~vino
Improving Data Management
-
Delta Compression
-
Delta/diffential data compression research at UC SAnta Cruz, and IBM Almaden.
http://csl.cse.ucsc.edu/projects/delta-compression/
http://www.almaden.ibm.com/cs/delta.html
-
Reducing I/O latency through improved file prefetching. Work conducted at the
Concurrent Systems Laboratory,
in the
Jack Baskin School of Engineering,
University of California, Santa Cruz.
http://csl.cse.ucsc.edu/projects/predictive-caching/
-
This project from UC Berkeley is attempting to build
"adaptive, highly-available, self-maintaining, self-tuning back-end servers for
storage-intensive network services"
http://iram.CS.Berkeley.EDU/istore/
-
Good source of analytical work on optimal performance in storage systems, in the context of the
Laboratory of Distributed Multimedia Information Systems and Applications of Crete
(MUSIC), in the
Technical University of Crete, Greece.
http://www.music.tuc.gr/Research/HERMES.htm
-
Optimised storage layouts for specific workloads, at the University of California, San Diego.
http://www-cse.ucsd.edu/groups/gemini/
-
Assistance for navigating large amounts of information through
"desktop repositories that archive both
base content and user-specific meta-information."
Adaptive Personalized
Information Retrieval from MIT's LCS.
http://haystack.lcs.mit.edu/
-
Provides flexible associative access to the system's
contents by automatically extracting attributes from files.
Programming Systems Research Group at MIT.
http://www.psrg.lcs.mit.edu/publications.html#sfs
-
A stackable interface file system.
http://www.cs.columbia.edu/~ezk/research/fist
-
Work from MIT's LCS "to define Input/Output
Automaton specifications for key communication
and memory building blocks"
http://theory.lcs.mit.edu/tds/ftds.html
-
More work from MIT's LCS for the input/output automaton model.
http://larch-www.lcs.mit.edu:8001/~garland/ioaLanguage.html
New Storage Technologies
-
MRAM
-
Replace capacitance with a magnetic recording, and you have
a no-compromise NVRAM that uses less power than DRAM.
http://www.wirednews.com/wired/archive/8.04/mram.html
http://www.imec.be/mcp/nmc/magneto/mram/mram.htm
http://www.almaden.ibm.com/st/projects/magneto/mtj/
http://www.research.ibm.com/resources/magazine/1998/issue_3/cmos398.html
-
The following link is to another "MRAM" but this is a parallel 3-D memory based
on principles similar to an MRI. SO it's "Magnetic Resonant Address Memory" to be
precise.
http://www.winbmdo.com/scripts/sbir/firm.asp?firm_id=1395
-
Data storage using volume holographic techniques. "Data crystals" are not sci-fi anymore.
http://www.almaden.ibm.com/st/projects/holography/
-
Company offering solutions based on Holographic data storage. After developing a few prototypes that did fingerprint searches for security systems, this web site seems to have completely changed focus.
http://www.holoplex.com/
-
Dimitri Psaltis' optical information processing group. Conducting research on
holographic memories, optical correlation and pattern recognition, liquid crystal deflector and
modulators, and other projects. Of particular note is their work in high capacity holographic
stores.
http://optics.caltech.edu/
-
UCSC faculty involved in holographic data storage research.
http://www.cse.ucsc.edu/~claire/
-
Holgraphic data storage on a roll of adhesive tape.
http://www.eml.villa-bosch.de/english/research/optimem/optimem.html
-
Alternatives to regular magnetic recording from IBM Almaden
SST.
Extremely dense storage.
http://www.almaden.ibm.com/sst/html/newmech/newmech.htm
-
Another alternative to regular magnetic recording from
the makers of
Millipede, the
molecular abacus, the
molecular amplifier, and the
molecular wheel.
Non-magnetic and yet incredibly dense storage.
A short piece on Millipede can be found here.
http://www.zurich.ibm.com:80/st/storage/thermomech.html
-
Another alternative to regular magnetic recording from IBM Almaden.
Extremely dense storage, addresses superparamagnetic limit by
making single-grain magnetic bits.
http://www.almaden.ibm.com/st/projects/patternedmedia/
-
This group at the Research Laboratory of Electronics (RLE)
has demonstrated RF signal gain in a directly modulated fiber
transmission system, and have done excellent work relevant to
patterned media.
http://rleweb.mit.edu/groups/g-opt.htm
-
Micromachines for data storage.
http://www.lcs.ece.cmu.edu/research/MEMS/index.html
-
A company claiming near production of practical MEMS alternatives to current magnetic disks.
http://www.nanochip.com/
Network Attached Storage
-
NASD @ NSICC
-
The National Storage Industry Consortium (NSIC home for the
Network Attached Storage Devices (NASD) project.
http://www.nsic.org/nasd/
-
Carnegie Mellon University's
Network Attached Secure Disks (NASD) project.
http://www.pdl.cs.cmu.edu/NASD/
-
A real-time, low power, and fault tolerant Network Attached Storage
Device
http://www.ecsl.cs.sunysb.edu/phoenix.html
-
We really need a link or page for Ben's work ... we don't have one up yet, neither do IBM.
U
-
A presenttation on the need for commercial network attached peripherals (NAP)
for HPSS goals.
http://www.llnl.gov/liv_comp/siof/hpss_nap_wg.html
Miscellaneous
-
The Rio (RAM I/O) Project
-
The use of "reliable memory" for storage.
http://www.eecs.umich.edu/Rio/
-
Data dissemination in assymetric network environments (server to client
bandwidth much higher than client to server), using push-based
tehniques. Brown University.
http://www.cs.brown.edu/research/ddg/bdisks.html
-
A past University of Maryland project on
data dissemination in assymetric network environments
(investigating web-push).
http://www.cs.umd.edu/projects/bdisk/
-
Uses a small log disk, referred to as cache-disk, as a secondary disk cache
to optimize write performance.
http://www.ele.uri.edu/research/hpcl/DCD/DCD.html
-
Storage Area Networks at CMU.
http://www.ece.cmu.edu/~asn/
-
Examining causes of performance problems with the
buffer pool of the DB2 DBMS.
http://www.cs.usask.ca/research/research_groups/discus/db2/index.html
-
An implementation of the Solaris Doors API on the Linux operating system.
The filesystem as a universal namespace.
Formerly at Brown University.
http://www.rampant.org/doors/
-
An Analysis of the Ext2fs structure by Louis-Dominique Dubeau.
http://step.polymtl.ca/~ldd/ext2fs/ext2fs_toc.html
-
An article on digital libraries, with references to works on wide-area
file systems applied to WWW (like the new idrive and enterprise file system
work). Not too surprising .. written by Transarc researchers.
http://www.dlib.org/dlib/february98/bowman/02bowman.html
-
A paper from the San Diego Supercomputer Center.
http://www.sdsc.edu/NARA/Publications/OTHER/Persistent/Persistent.html
-
On-Line Analytical Processing of Large Multidimensioal Data Sets.
http://www.ece.nwu.edu/~sgoil/parsimony
-
A paper describing the Sawmill, the logging file system for a high performance
RAID array. Developed by Ken Shirriff
under the Sprite project at Berkeley.
http://sunsite.berkeley.edu/Dienst/UI/2.0/Describe/ncstrl.ucb/CSD-95-862
-
A deliberatley simple file system for the University of Maryland's
Roadrunner
mobile OS.
http://www.cs.umd.edu/~fwmiller/roadrunner/fs.html
-
Part of Stanford's FLASH project.
An OS for a the FLASH multiprocessor, and research work on parallel performance
issues (including networks of workstations).
http://www-flash.stanford.edu/OS/
Last updated Thursday, 06-Sep-2001 02:24:15 PDT by A. Amer (amer4@cse.ucsc.edu)
